Finger Twitches are More Frequent in REM Sleep Than in Non-REM Sleep

Introduction: Abnormal rapid eye movement (REM) sleep is often symptomatic of chronic disorders, however polysomnography, the gold standard method to measure REM sleep, is expensive and often impractical. Attempts to develop cost-effective ambulatory systems to measure REM sleep have had limited success. As elevated twitching is often observed during REM sleep in some distal muscles, the aim of this study was to assess the potential for a finger-mounted device to measure finger twitches, and thereby differentiate periods of REM and non-REM (NREM) sleep. Methods: One night of sleep data was collected by polysomnography from each of 18 (3f, 15m) healthy adults aged 23.2 ± 3.3 (mean ± SD) years. Finger movement was detected using a piezo-electric limb sensor taped to the index finger of each participant. Finger twitch densities were calculated for each stage of sleep. Results: Finger twitch density was greater in REM than in NREM sleep ( p < 0.001). Each sleep stage had a unique finger twitch density, except for REM and stage N1 sleep which were similar. Finger twitch density was greater in late REM than in early REM sleep ( p = 0.005), and there was a time–state interaction: the difference between finger twitch densities in REM and NREM sleep was greater in late sleep than in early sleep ( p = 0.007). Conclusion: Finger twitching is more frequent in REM sleep than in NREM sleep and becomes more distinguishable as sleep progresses. Finger twitches appear to be too infrequent to make definitive 30-second epoch determinations of sleep stage. However, an algorithm informed by measures of finger twitch density has the potential to detect periods of REM sleep and provide estimates of total REM sleep time and percentage. Keywords: finger, twitch, REM sleep, actigraphy, wearable